brainstorm

name	brainstorm
description	Explore an unclear raw request and converge it into one operator-locked Decision-Complete Work Brief 2.0.0.
vibe-protocol	work-brief-producer
vibe-input-artifact	raw-intent
vibe-output-artifact	work-brief
runtimeExecutionClaim	pending-live

brainstorm

Explore an unclear raw request and converge it into exactly one operator-locked Decision-Complete Work Brief 2.0.0.

What this skill owns

One Work Brief 2.0.0 under .vibe/work/<work-id>/work-brief.json. It is the only artifact this skill may produce; it carries the decision register, rejected alternatives, coverage with omission rationales, the exact Project Atlas closure cited by digest, research/prototype/diagnosis refs, acceptance criteria with observable seams and proof classes, capability requirements, the discovery review, and the operator lock. It may not leave implementation-affecting unknowns at decision_complete and may not produce an implementation plan, build result, or ship packet.

Protocol

1. Capture the raw request under .vibe/work/<work-id>/raw-intent.md before normalizing it.
2. Read the exact Project Atlas closure (root pair, then the smallest relevant topic set); record document IDs, paths, statuses, and digests in the brief. A roadmap item or likely statement is never plan/build authority.
3. Map the applicable decision coverage dimensions breadth-first; explain every omitted dimension. Verify repository facts directly; use research for external facts and prototype where prose cannot settle a material choice.
4. Ask one operator decision at a time, each with a recommendation, rationale, alternatives, consequences, and a reopening condition. Record every resolved decision immediately in the decision register; facts are looked up, not delegated to the operator.
5. Run the six-axis discovery review (completeness, contradiction, research, scenario, implementation-question simulator, scope). Any unresolved implementation-affecting question blocks lock.
6. Present the exact persisted Work Brief 2.0.0 to the operator and record the operator lock binding approver identity/time to the effective artifact digest.

Completion

The brief is decision_complete only when implementation-affecting assumptions/openQuestions are empty, the discovery review carries zero unresolved questions, and the operator lock is recorded. A short or vague request stays sparse until real decisions close it — never padded into a fake decision_complete.

Authority and recovery

The one project skill is the sole Project Atlas entrypoint; foundation enrichment routes to project, not here. A concrete task request routes here. Recovery is always forward-only inside this run. A miss is never a dead-end rejection and never instructs the operator to re-invoke an earlier skill. The current run launches the forward-only gap resolver, persists a scoped Decision or Plan Amendment bound to the exact amendment digest, revalidates affected closure with targeted invalidation of only the impacted evidence, and resumes automatically. When bounded in-run resolution cannot establish a safe path, the run records an honest resumable wait or hard block with an exact next action — it never converts that into prose-green and never bounces the operator back to an earlier skill as recovery.

Artifact carrier

• Persist this skill's one output under .vibe/work/<work-id>/ as a UTF-8 JSON or Markdown artifact; chat history alone is never a carrier.
• Verification evidence lives under .vibe/evidence/<work-id>/; the starter runner catalog is .vibe/registry/runner-catalog.json.
• Runtime execution claim: pending-live — native pi loading and execution stay unproven until a recorded authenticated runtime witness exists.

Forbidden outputs

This skill never produces: implementation-plan, build-result, ship-packet, push, pull-request. Recovery stays forward-only inside this run; a miss is resolved in place through the current-run gap resolver, never by handing the operator back to an earlier skill.